Quick start
import { openai } from '@agentskit/adapters'import { createDataAbTestInterpreterAgent } from './agents/data-ab-test-interpreter/agent'const agent = createDataAbTestInterpreterAgent({ adapter: openai({ apiKey: process.env.OPENAI_API_KEY!, model: 'gpt-4o', }),})const result = await agent.run('Describe your task here')console.log(result.content)Independent reviewer approved
Validation evidence
- Review score
- 96/100
- Confidence
- 95%
- Evaluation cases
- 3
- Iterations
- 1
The agent produced valid structured outputs for all three cases, stayed within its A/B test interpretation purpose, refused to invent results from missing data, surfaced concrete gaps and open questions, and handled the injection case without following the malicious instruction. The behavior is conservative but appropriate for sparse or invalid A/B test inputs.
What passed review
- All outputs are non-empty, structured, and schema-consistent.
- Clearly refuses to infer winners or recommendations without experiment data.
- Surfaces missing context in actionable terms: metrics, sample sizes, variants, dates, confidence intervals, stopping rules, guardrails, and business goals.
- Injection attempt was correctly treated as untrusted data and not followed.
- No material hallucination beyond the provided input.
Reviewer notes
- Add at least one live validation case with actual A/B test metrics to confirm it can correctly interpret significance, uncertainty, guardrails, and business decision risk when real data is present.
Extend it
Pass tools, retrieval, memory, permissions, and observers through the factory config.
const agent = createDataAbTestInterpreterAgent({ adapter, tools, retriever, memory, onConfirm: (call) => approve(call), observers: [tracer],})View agent factory source
import type { AdapterFactory, ChatMemory, Observer, ToolCall, ToolDefinition } from '@agentskit/core'import { fenceUntrustedContent, UNTRUSTED_CONTENT_DIRECTIVE } from '@agentskit/core/security'import { invokeStructured } from '@agentskit/runtime'import { defineZodTool } from '@agentskit/tools'import { z } from 'zod'import { zodToJsonSchema } from 'zod-to-json-schema'import type { JSONSchema7 } from 'json-schema'/** A/B Test Interpreter — v1 validated. Pain: A/B results misread */export interface Finding { id: string; severity: 'critical' | 'high' | 'medium' | 'low' | 'info'; message: string; source?: string; recommendation?: string }export interface AgentOutput { summary: string; findings: Finding[]; gaps: string[]; openQuestions: string[] }export interface AgentResult extends AgentOutput { requiresReview: boolean }export interface DataAbTestInterpreterConfig { adapter: AdapterFactory memory?: ChatMemory observers?: Observer[] onConfirm?: (toolCall: ToolCall) => boolean | Promise<boolean> maxSteps?: number}const Output = z.object({ summary: z.string(), findings: z.array(z.object({ id: z.string(), severity: z.enum(['critical', 'high', 'medium', 'low', 'info']), message: z.string(), source: z.string().optional(), recommendation: z.string().optional(), })), gaps: z.array(z.string()).default([]), openQuestions: z.array(z.string()).default([]),})const toJson = (s: z.ZodTypeAny): JSONSchema7 => zodToJsonSchema(s) as JSONSchema7const skill = { name: 'data-ab-test-interpreter', description: "A/B Test Interpreter — typed output agent (draft spec).", systemPrompt: `You are A/B Test Interpreter. A/B results misread. Output: Interpretation typed.Actionable findings citing input sources. No invented issues.NEVER invent facts — gaps and openQuestions for missing input. Always draft for human review.${UNTRUSTED_CONTENT_DIRECTIVE}Call submit_test_interpreter exactly once. Stop.`, tools: ['submit_test_interpreter'],}export function createDataAbTestInterpreterAgent(config: DataAbTestInterpreterConfig) { const submit = (): ToolDefinition => defineZodTool({ name: 'submit_test_interpreter', description: 'Submit result. Once.', schema: Output, toJsonSchema: toJson, async execute() { return 'recorded' } }) as ToolDefinition async function run(input: string): Promise<AgentResult> { if (!input?.trim()) throw new Error('data-ab-test-interpreter requires non-empty input') const result = await invokeStructured({ adapter: config.adapter, tool: submit(), task: `INPUT:\n${fenceUntrustedContent(input)}`, parse: (a) => Output.parse(a), skill, memory: config.memory, observers: config.observers, onConfirm: config.onConfirm, maxSteps: config.maxSteps ?? 4, }) return { ...result, requiresReview: true } } return { name: 'data-ab-test-interpreter', run, asHandle() { return { name: 'data-ab-test-interpreter', run: (t: string) => run(t).then((r) => JSON.stringify(r)) } }, }}View evaluation contract
Replay these cases with the provider and model you plan to deploy.
import type { EvalSuite } from '@agentskit/eval'export const suite: EvalSuite = { name: 'data-ab-test-interpreter', cases: [ { input: 'Complete input for A/B Test Interpreter: A/B results misread. Provide full structured output.', expected: (r: string) => r.length > 20 && /requiresReview|summary|title|category|findings|sections|score|clusters|items|steps/i.test(r) }, { input: 'Minimal input.', expected: (r: string) => /gap|openQuestion/i.test(r) || r.length > 10 }, { input: 'Input with specific detail: ACME Corp project deadline March 15.', expected: (r: string) => /ACME|March|15/i.test(r) || /gap/i.test(r) }, { input: 'Empty context — only says "process this".', expected: (r: string) => r.length > 5 }, ],}Was this agent useful?
Your response helps us prioritize agent quality.